Here I'm again asking for your help. I start migrating users from exchange 2010 to 2016. I prepared the EV server month ago for this migration and add some user already at exchange 2016, after migrating 1000 user we noticed that took 2 hours to archive 50 mailboxes (747 items - 116,6Mb). Some EV servers against exchange 2010 in the same 2 hours archive 2085 mailboxes (18047 - 7,2 GB), for example, running archiving tasks in report mode against 1 mailbox in exchange 2016 takes 20 min in exchange 2010 is less than a minute.
I found this article but no luck
Dtrace indicate all message are processed in the same milisecond
Dtrace show that each folder is processed under a 1 sec
Enterprise Vault and Exchange servers don't have any error in the Event Viewer.
It's look like the Vault Service Account is throttled, I checked and seems ok.
I'm still in Enterprise Vault 11.0.1 CHF5, I'll upgrade to EV 12 after exchange migration terminates
I'll run the powershell script SetEVThrottlingPolicy.ps1 just to be safe.
Running the PS is a good step, just to be sure. There is a way to check if it is indeed set on/for the new Exchange server using adsieedit, but I cannot locate the document right now.
It might also be necessary to move the VSA mailbox to Exchange 2016. I recall having seen forum entries about the performance drop earlier, but also those, I cannot find.
If I recall correct, there was an issue with archiving a Journal Mailbox, and you had to run a script to create "EV View" mailboxes in the Journal Mailbox. That might or might not apply here too.
Your best option (again) is to call support. DO mention you plan to upgrade to EV12.5.3 as soon as you can. Otherwise they will deny assistance, due to you using EV11.x. Even a statement like 'that is known, and fixed in 12' should help already right :)
I believe I found the problem. Because I didn't not found any error, I did the following
So I bypass the Exchange load balancer by adding an entry in the EV server hosts file. Everything is working.
I'm waiting for the Networking Team to identify why the load balancer is throttling the connection.
After a few backs and forward with the network team we got them to turn off Nagel's algorithm for our server and the problem was fixed. Just follow this article https://www.veritas.com/support/en_US/article.100032406 .